If a matrix has a full set of independent eigenvectors, something wonderful happens: in the basis made of those eigenvectors, the matrix acts as a plain diagonal stretch. We can write

where the columns of P are the eigenvectors and D is the diagonal matrix of eigenvalues. Read right to left: P^{-1} rewrites a vector in eigen-coordinates, D stretches along each eigen-axis, and P translates back. The tangled matrix was a simple stretch all along — just along the wrong-looking axes.

A stretch along its own axes

The faint grid is aligned to the eigenvectors of A = \begin{bmatrix} 2 & 1 \\ 1 & 2 \end{bmatrix}. Drag the slider to apply A: the grid doesn't rotate or twist — each eigen-axis simply stretches by its eigenvalue (×3 along one, ×1 along the other). In its own basis, A is nothing but two scalings.

Why diagonal is gold

Diagonal matrices are trivial to raise to powers — and so are diagonalizable ones, because A^n = P D^n P^{-1} just powers the eigenvalues. That single trick evaluates matrix powers instantly, solves systems of differential equations, computes the long-run behaviour of Markov processes, and sets up the decompositions that data science leans on. Eigenvectors turn "apply this matrix a thousand times" into "raise two numbers to the thousandth power."